Modular Analysis Of Cancer Gene Expression Networks

Cancer,or malignant tumor,has now become one of the major global public health problems that threaten human life and health.How to detect cancer early and prevent and treat cancer effectively is a problem that needs to be solved urgently.The molecular mechanism of cancer is still unclear,and most transcriptomics studies on cancer have focused on the differential expression of each gene between tumors and normal tissues,the other disturbances caused by cancer including gene mutations,the change of gene modules,which may be the strength crucial for the cancer diagnosis,treatment and prognosis,are more or less ignored.Among many cancers,breast cancer is a complex disease that is generally prevalent among women and,it’s one of the most common malignant tumor among women.Since the late 1970s,the incidence of breast cancer has been rising globally,and there has been a trend toward youth.There are about 1.671 million new cases of breast cancer worldwide each year,and about 522,000 people die of breast cancer every year.In this thesis,we take breast cancer as an example and conducts research from multiple perspectives.We hope that this research can provide a complete framework for the study of cancer and be helpful for the diagnosis,treatment and prognosis of cancer.We download the RNA expression profiles of 1109 breast cancer tissues and 113 non-tumor tissues from the Cancer Genome Atlas(TCGA)database,as well as the clinical information of breast cancer patients.We carry out a complete study framework of breast cancer from the following aspects:Firstly,we use DESeq2 for differential expression analysis to identify 14118 differential genes such as MMP11 and COL10A1,and we perform survival analysis on these differential genes to find genes that affect the survival time of breast cancer patients such as UBE2T.Secondly,the weighted gene co-expression network analysis(WGCNA)was used to construct a gene co-expression network.In order to compare the module difference between the tumor and normal tissues,the module difference connectivity(MDC)analysis was performed,and 23 modules was found in the tumor network having significantly different co-expression mode.The gene function enrichment analysis of these differential modules,and the Go terms related to breast cancer,such as MHC protein complex,leukocyte activation,regulation of defense response,lymphocyte activation,etc.,have enriched in the top differentiation modules.In addition,in order to study the expression of genes in tumor and normal states at the same time,a bi-Gaussian mixed model was used to cluster genes,and gene set association analysis(GSAASeqSP)was performed.